#1785

[@absurdengineering]

These supplies are little engineering gems. A couple transistors and you get such a well-behaved power supply. I’m a fan of those, got most of the models they made.

[@bayareapianist]

Infact, this was the first circuit I learned in elementary school (maybe 5th) in 70s. I wanted to have a power supply for my radio. But a simple PS had too much hoom and noise. Then someone sat and tought me the basics of transistors. I had already learned the Ohm's low the year before. Later, I wanted to use NiCd batteries and needed to limit the current once charged. Then I was guided to add a current limit and control the main transistor. I remembered I had a zener diode connected to the base and ground. It was just like yesterday and these simple circuits made my mind to become an EE at the age of 12.

[@StickySli]

I started to work about a month ago at an Industrial and Control Engineering laboratory at my university and now I can see these circuits with awe. It's incredible the though these engineers have put to build these devices in an era of pens and pencils. I used to have some old books that smelled amazing with hand-drawn schematics like the ones you present in this video. I loved to read it when I was little. Good stuff!

[@elektronikvideos-bremen2873]

There a some more things to say: 1. Vbe of Q4 compensates for the tempco of the pass transistors 2. The Zeners are necessary to create a bypass for Q5 which pulls the emitter of Q4 below the base potential of the pass transistors to turn them off.

[@CannonballCircuit]

thanks for adding more insight! Would you be able to elaborate on point #2? Is it saying that if there were no reference voltage created by the Zeners, Q5 wouldn't be able to turn on Q4?

[@mnoxman]

On all three of my HP/Harris supplies the main big caps are like 150/160Volt even though they maximum voltage on one is 70 volts. One of them was like 45 years old and the main cap was good when I re-capped it for the other caps. So I am no longer in the camp of "to 'reform' the cap you must be at rated voltage'. The item you glossed over with the negative voltage is that is how the supply can go from 0 volts to N volts of the supply. Most supplies would only go from +0.6 to +1.4 volts to N. By referencing all of the semiconductors to a ground of -6 volts the supply can be "calibrated" to work from -0.6 volts to +N volts/amps. They also reference everything to the iron core for ground. As you mentioned none of the secondary windings have a center tap but they use that "core ground" to be the center tap. This lets it have the secondary windings "float".

[@nagyandras8857]

indeed in practice if a given circuit should never be able to produce say... 12 volts, and you need a cap in that circuit, its never ever a good idea to select say.. a 16v rated one. if you can afford the space/cost/whatever , go higher. a lot higher. having a rating of least twice as much you should ever see from the circuit is actually a good practice. applies to replacing caps, if it fits you can not possibly go wrong with something rated for a higher voltage than the original. same applies to the rated heat they can take.

[@mnoxman]

@@nagyandras8857 There is a lot of "mythical talk" about voltage rating and "reforming" and "in circuit performance". The fact that my three supplies had a 160v cap when the max the 70v one would see is 100Vp. It operated for 45 years. For me, ended the voltage rating "voodoo gossip". I won't put a 3.3v cap in a 25 vol circuit but a 160v cap in a 5v circuit my care factor is less. I had two more that are the 30 volt ones and the same size cap was in those as well so the voltage rating delta was far more in those.

[@Stevie_D]

Excellent explaination - thank you!

[@ujjobbagya]

Wow, thanks for this!

[@ivanpopovic9503]

Though this, and 'floating' regulation are quite known and used solutions, made famous by Harrison labs/HP, there is one circuit by Harrison lab, used in their SCR regulated or preregulated power supplies that is simply stroke of genius. It is their firing angle controll/scr trigger circuit with a single transistor in a monostable blocking oscillator configuration. This circuit has some really tricky parts, solved exclusively with passive components and a few diodes, and here I'm refering to zero cross detection, and circuit for disabling retriggering of a blocking... I do not know the name of an engineer who designed it, but that guy must have had IQ of 3000 at least. 😂 I mean, to design circuit this simple yet with this much functions one must be kinda mad scinetist!

[@IMSAIGuy]

Yes, I've seen that pre-regulation SCR circuit in a different power supply I have. Quite clever.

[@yoctometric]

Great video! Would you consider using a narrower brush when you are marking up the circuits? It was a bit hard to read them.

[@IMSAIGuy]

yes, that was my first time using that tool. I agree

[@eigenvector7035]

This is an excellent video - what took me a while to figure out was the return current path to the virtual voltage source ("battery") of that auxiliary voltage. Am I correct it goes through the BE junction of the power transistor pair?

[@Georgesch3]

Great explanation, and very clearly delivered. Reminds me of the storied Barney Oliver Audio Amplifier. Maybe you can touch on that bit of HP history sometime. Thanks! 👍

[@IMSAIGuy]

Both Barney and his amp predated my time a HP Laboratories I never had a chance to hear that amp in action. It sure helped that HP had its own transformer winding factory in Palo Alto.

[@robinbrowne5419]

It's great to be retired. I get to sleep in, eat honey flavoured shreddies for breakfast and watch KZbin videos 👍

[@thushararathnayake]

Hi Imsai Guy, I have a problem with how to design an auto frequency sensing, anti-howling circuit for a condenser microphone pre-amp. Could you please help me to do it with one of your videos?

[@GaijinPaul]

Reminds me of how trigger transformers allow using N-channel high-side switching in H bridges. 4 of 8 40A IGBTs exploded in a plasma cutter power supply H bridge I'm sorta trying to fix.

[@marklundeberg7006]

And to avoid using a full bridge rectifier, they use a center tap on the floating winding, which saves 2 diodes. That seems like a really uneconomical decision these days. (Unless you have limited transformer selection... then the center tap option yields half the voltage which may be more appropriate).

[@robballantyne3]

On the face of it the "battery" winding looks like it's applying near to 26v Vbe of the pair of pass transistors. I see R34, which looks to be a 2w 10ohm resistor that would/could drop the voltage. It just seems like a lot of voltage applied to the bases to me.

[@PileOfEmptyTapes]

It's actually an 820 ohm. Basically the DC voltage stacked on top of the regulator output saved them a constant current source (for higher error amplifier gain and better PSRR) and keeps the regulator operating up to very low dropout. It's essentially a form a bootstrapping that's effective down to DC.

[@bob_mosavo]

Thanks 👍

[@jjoeygold]

Lends itself to Ltspice model and simulation

[@gorak9000]

that's usually what I do when I'm looking at some analog circuit and I'm lost as to what it's actually doing - throw it in spice!

[@meh583]

Can we get a link to the schematic?

[@mikesradiorepair]

Looks "almost" exactly like a Harrison 6200B power supply circuit. Only real differences I see is the 6200B has two voltage ranges (20/40), has a meter drive circuit and uses a full wave bridge rectifier. Otherwise looks exactly the same including the pass regulator transistor bias circuit.

[@IMSAIGuy]

www.keysight.com/us/en/assets/9018-03099/user-manuals/9018-03099.pdf

[@sergepetrov8598]

hparchive.com/Manuals/HP-6200B-Scheamtics.pdf

[@vckrijt]

Hej you sound different ... ;)